It can be challenging to find or locate a specific person or suspect from a large number of video surveillance cameras over any specified or extended time period. Existing approaches include manually searching through videos. However, even when individual people can be extracted from each camera's view, the user would need to consider each person in each camera over the time period to find the matching person. Such approaches are prohibitively time consuming, and search criteria are not always accurately known or executed.